Analysis on factors affecting blood transfusion volume of obstetric patients in a tertiary hospital in Beijing
10.13303/j.cjbt.issn.1004-549x.2026.08.004
- VernacularTitle:产科患者输血量影响因素分析
- Author:
Lihui FU
1
;
Chunya MA
1
;
Yuanyuan LUO
1
;
Xiaozhen GUAN
1
;
Shengfei TAI
1
;
Hongmei SHI
1
;
Yang YU
1
Author Information
1. Department of Transfusion Medicine, The First Medical Center of Chinese PLA General Hospital, Beijing 100853, China
- Publication Type:Journal Article
- Keywords:
blood transfusion volume in obstetric patients;
linear regression analysis;
logistic regression analysis;
random forest
- From:
Chinese Journal of Blood Transfusion
2026;39(8):1018-1025
- CountryChina
- Language:Chinese
-
Abstract:
Objective: To explore the influencing factors of blood transfusion volume in obstetric patients based on the national health development strategy, and to provide evidence for clinical transfusion management and maternal and infant safety. Methods: A total of 1 203 obstetric patients who received blood transfusion in a tertiary hospital in Beijing from 2013 to 2023 were enrolled retrospectively. Seventeen indicators including age, body mass index (BMI), hemoglobin and prothrombin time were selected as research variables. Descriptive analysis was conducted firstly. Scatter plot matrix was used to analyze the correlation between continuous variables and transfusion volume, and box plots were used to analyze the relationship between categorical variables and transfusion volume. Linear regression, logistic regression and random forest models were established for empirical analysis, and ROC curves and AUC values were compared to evaluate model performance. Results: Descriptive analysis showed that the distribution of obstetric blood transfusion volume was right-skewed, and multiple indicators were significantly correlated with transfusion volume. Linear regression indicated that 9 variables such as age, BMI and hemoglobin had linear correlation with transfusion volume. Logistic regression confirmed that 7 variables including BMI and hemoglobin were statistically correlated with the probability of massive blood transfusion. Random forest model screened the core influencing factors, which were ranked by importance: hemoglobin, prothrombin time, hematocrit, BMI, activated partial thromboplastin time, platelet count and age. The AUC values of logistic regression and random forest were 0.78 and 0.81 respectively, and the latter had better predictive performance. Conclusion: Prenatal anemia, coagulation disorders, advanced age, obesity, gestational hypertension, placental abnormalities and organ damage can increase the risk of massive blood transfusion in obstetric patients. The random forest model has better analytical and predictive ability in this study. The results can provide reference for accurate blood transfusion and risk prediction in obstetrics.
